It is well recognized that current batteries of genetic toxicology assays exhibit relatively high sensitivity, meaning they effectively identify genotoxic carcinogens. However, a critical deficiency with current approaches exists-namely, the specificity of the in vitro mammalian cell genotoxicity tests is low, as they yield a high incidenc of positive results that do not have in vivo relevance (so-called misleading or irrelevant positives). This high incidence of irrelevant in vitro positive results leads to extensive and costy additional testing, often with whole animal models, or else abandonment of potentially valuable products. We will address this major problem with current in vitro mammalian cell genetic toxicity assays by developing commercial kits that enable an automated testing strategy that exhibits both high sensitivity and specificity. This system will categorize positive results according to the predominant mode of genotoxic activity, and importantly, will reliably identify irrelevant mode(s) of action that are likely to be nonoperational in vivo. A secondary objective of the proposed work is the development of methods for characterizing clastogens. That is, we will develop tools that elucidate whether clastogenic activity is the result of DNA-reactivity, or whether it is mediated by an indirect mechanism.